CNCF AI Landscape¶

KFabrik is part of the broader Cloud Native Computing Foundation (CNCF) ecosystem focused on AI and machine learning workloads. This page provides context on related projects and how they work together.

Model Serving & Inference¶

KServe
Kubernetes-native model serving platform for machine learning models. Provides serverless inference with autoscaling, canary deployments, and multi-framework support.

vLLM
High-throughput and memory-efficient inference engine for large language models. Optimized for serving LLMs at scale with advanced batching techniques.

Orchestration & Management¶

Karmada
Multi-cluster Kubernetes management system that enables workload distribution across multiple clusters, useful for distributed AI training and inference.

LLMD
Kubernetes operator for managing large language model deployments and lifecycle operations.

How KFabrik Fits In¶

KFabrik complements these projects by:

Simplifying deployment of AI workloads across the CNCF ecosystem
Providing abstractions that work with KServe, vLLM, and other inference engines
Enabling multi-cluster AI workflows through integration with Karmada
Democratizing access to cloud-native AI tools for developers

Integration Patterns¶

With KServe¶

KFabrik can deploy models that are served by KServe, providing a simplified interface for model deployment and management.

With vLLM¶

Use KFabrik to orchestrate vLLM deployments for high-performance LLM serving across multiple environments.

With Karmada¶

Leverage Karmada's multi-cluster capabilities to distribute KFabrik-managed AI workloads across different regions or cloud providers.